وظيفة مهندس مبادرات AI لدى Innovative Systems (ISys) في الرياض

Company Description

Innovative Systems (ISys) delivers technology solutions tailored to the specific needs and requirements of each customer, with a focus on value, performance, and strong support services. Based in Riyadh, Saudi Arabia, ISys works closely with a global network of technology and business partners to provide access to cutting-edge tools and platforms. The company emphasizes customized, high-impact technology implementations rather than one-size-fits-all products. Team members collaborate with international partners and clients, gaining exposure to diverse industries and technology ecosystems. ISys offers a dynamic environment for professionals interested in shaping advanced, customer-centric solutions.

Role Description

You'll own our RAG capability end-to-end - from ingesting technical documents (manuals, wiring diagrams, project specs, compliance standards) through to delivering accurate, trusted answers to our engineers, project managers, and field staff.

Working directly with our AI Initiative Engineer, you'll move fast: prototype, evaluate, iterate, and ship. This is a greenfield technical challenge in a complex, document-rich industrial environment.

Responsibilities

• Design and build RAG pipelines - ingestion, chunking, embedding, vector indexing, retrieval, reranking, and LLM response generation.
• Select and optimize vector databases (Pinecone, Weaviate, Qdrant, pgvector) for accuracy, latency, and cost.
• Implement advanced retrieval strategies: hybrid search, HyDE, multi-query, contextual compression.
• Integrate LLMs via LangChain or LlamaIndex (OpenAI, Claude, Mistral, or open-source alternatives).
• Build document ingestion pipelines for PDFs, Word docs, CAD files, scanned records, and structured data.
• Evaluate and tune quality using RAGAS or custom benchmarks - track faithfulness, relevance, and recall.
• Expose RAG capabilities via APIs and chat interfaces to internal tools, ERP, and field apps.
• Keep knowledge bases current, accurate, and governed - define ingestion standards and update cadences.

Qualifications

• Experience: Mid-Senior (5-7+ years)
• 3+ years in NLP, LLM engineering, or applied ML - with at least one production RAG deployment under your belt.
• Strong Python and experience with LangChain and/or LlamaIndex.
• Hands-on vector database experience and a solid grasp of embedding models (BGE, Sentence Transformers, OpenAI, Cohere).
• Familiarity with hybrid retrieval (BM25 + dense) and reranking techniques.
• Comfort with cloud deployments (Azure OpenAI, AWS Bedrock, or GCP Vertex AI) and Docker/CI-CD basics.
• A systematic approach to debugging hallucinations, retrieval failures, and context mismatches.
• Bonus: knowledge graphs, multimodal RAG, LLM fine-tuning, or experience in engineering/industrial environments.